Equipment Design: Mechanical Aspects

Prof. Shabina Khanam

Department of Chemical Engineering
Indian Institute of Technology – Roorkee

Lecture 07
Design of Heads

Welcome to the second lecture of week 2 and here we are discussing design of heads. This topic
we have started from the previous lecture where we have discussed different types of heads and in
this particular lecture, we will discuss procedure to design these heads. As flat head is the simplest
one, we will start design from that and then we will proceed towards conical head and then for
dished heads okay. So, let us start with design of flat heads.
(Refer Slide Time: 00:59)

Now, in this case, we will calculate thickness of flat unstayed circular heads and curves, okay?
So, you see here we have one term that is unstayed circular heads, what is the meaning of unstayed
circular head? To illustrate this or to make you understand, I am taking the example of this glass,
now if you see this glass, this is open part of this glass where I need to cover, so what is the meaning
of unstayed. To understand this, let us first discuss what is stayed.

So, stayed means, when this is the opening which I need to cover, so what I do, I put some of the
metal strip and those strips will be welded at its edges. For example, I am taking one strip of this
much diameter, I hope you are getting this, one section will be welded over here, another would
be welded over here. Similarly another strip of same diameter and another strip diameter and this
way, I will put different strips and over this the head will be stayed. So, it basically gives more
strong positioning of head, but here we are discussing unstayed where this type of structure or this
type of support is not available to the head.

Even that type of structure can be used in the shell also to keep its geometry regular. For example,
if I am having this glass and this is the whole length of the shell, so that type of the structure, it
means when we weld the strip at two edges, when we keep the weld of those strips equal to the
diameter of shell, so it means when we weld it, it will keep its shape circular only, okay? I hope
you are getting it.

So that type of structure is usually used in pressure vessel when it is having larger diameter, but it
does not interrupt the process because in between it has open space for fluid to move, okay. So,
here, as far as design is concerned for flat head, we are considering design of unstayed flat heads.
So, to calculate thickness of flat head, we can use the following expression, as you see, this is the
expression which we will use. Here t is the minimum thickness of flat head and that is equal to
𝐶𝐷 𝑝
10
√𝑓 where p is the design pressure in kgf/cm2 and f is the allowable stress in kgf/mm2 and due

to this here we have the term 10.

If I am considering pressure and allowable stress in same unit, that is N/m2, then this 10 will not
appear. In this formula, t which is the minimal thickness is found mm and C is the factor that
depends on the method of attachment to the shell. It means how this head is attached to the shell,
accordingly value of C varies okay. And further D is the diameter or short span measured. It
means D is not always the diameter, you can consider this as an effective diameter, which will
depend on how head is attached to the shell.

𝑝
Here I am having this expression where 𝐶𝐷𝑒 √𝑓 where both pressure and allowable stress are N/m2,

so in books you can find this expression. This expression is available in IS: 2825-1969 code. So,
here in book instead of D it is De is mentioned, where it is called as effective diameter, C is the
factor. Depending upon different attachments, now we will see how to find out the value of factor
C as well as D that is the effective diameter. So, let us start that with flange flat head butt welded
to the vessel. For this connection, we will see the value of factor C as well as diameter.

Now, if this is the image where you consider this particular section is resembling to the shell and
over the shell, head is attached. So, if I consider placement of this, that placement is horizontal,
okay. So this head is attached to the horizontal shell, okay. ts is the standard thickness of shell and
this t is the thickness of flat head which we need to find and here you see, in this particular section,
this is basically the curved part of formed section. This is not flat section.

This is flat formed section, okay? In that case, D will be Di. Di is what? Di basically inner
diameter of head, okay, and C is the factor which has the value 0.45. So, when we have flange flat
head butt welded to the vessel, we can consider C and D in this way and these values we can put
in this expression to calculate the value of thickness.
(Refer Slide Time: 06:53)

Now, further, we have another attachment or another connection, and that is plates welded to the
inside of the vessel, so how it looks like? Here, I am having this shell and inside this shell head is
attached. If you consider the previous diagram, there head is attached to the shell through butt
welding. I hope you remember different types of welding which we have discussed in lecture 4 of
week 1 in weld joint efficiency factor, okay.
So, here we have this flat head which is attached inside to the shell, here also and in this diagram
also. Now what we need to focus on that if you consider this section, okay, why this section is so
black, okay. This section is so black because when we are welding it, the material through which
it is welded, like iron, that is filled inside the corner and when that filling is higher, we call that as
fillet. So, this black spot in this diagram as well as in this diagram, these are nothing but fillet us.
It has a particular thickness as 0.7𝑡𝑠 .

If it is more thick, it will be called fillet, otherwise that will be simply looking like a joint, okay.
So, when this type of connection I am having, D should be equal to Di and C value should be
greater than 0.55. Now, if you compare this image with the previous one, previous figure has
formed section. So, there C value was 0.45, but in this particular case, formed section is not there,
flat head is directly attached, so C will have greater value than 0.45 and if you remember the
expression for thickness, thickness is directly proportional to C.

So, when C value will be small, t will also be small, so in formed head we can have less thickness
in comparison to flat head as we are discussing now. Next, we have is the plate welded to end of
the shell. For example, if it attached like this, like this is the shell and here we have this attachment
of head, so if head is welded at the end of shell, in that case D would be Di and C would be 0.7,
okay. Now, if you compare this image as well as this image, in this image, I am not having any
fillet.

However, in this case fillet is made and that fillet will give slight curvature to this sharp corner and
here stress distribution will be more uniform in comparison to this figure and therefore C value
would be 0.7 over here, but here minimum C value would be 0.55. So, if it can work with 0.55,
keep that value, equal to 0.55 only.
(Refer Slide Time: 10:18)
Now next connection I am having is plate welded to the end of the shell with additional fillet weld
on inside, okay. So, this is noting but the previous diagram only. At the end of shell, head is
attached with fillet. So, you can see the value of C and D would be D in this case. Now, next I
am having is curves riveted or bolted with full face gasket to shells or flanges.

So, this connection is looking like this, where if I am having this flange section, okay. Flange is
basically used to joint two parts having same diameter. In detail, the flange design will be
discussed in week 3, okay, but for the time being you can understand that it will be used to connect
pipe of same diameter. Here, in this diagram you see, this is one section of flange and over this
we have flat head, so this flat head is connected to the flange, now here you see full face gasket.

Full face gasket means, this whole surface of the flange is covered with gasket, so if I am having
circular pipe, this particular section is placed around the pipe and this section will also be circular.
So, circular gasket of this much diameter will be kept over here and therefore it is called full face.
Now, if I need to attach this flange with flat head, here these 2 assemblies connected through bolt,
so one bolt is available here, another is available here and around the periphery different bolts are
available and it gives the diameter that we are showing as D in this diagram and which is called as
bolt circle diameter.

Now, if this is the case, D would be D as it is shown here and C would be 0.42. Further, if we have
covers with narrow face bolted flange joint, that is gasket is placed within the bolt circle, now how
it will look like, if you see, here it is horizontally placed, here we have demonstrated vertically
placed head with the flange. Here head is placed with the flange in horizontal manner. Now, if
you see this particular section, this is basically the gasket which is placed within the bolt circle.

Bolt circle, one is this and second edge will be up to here, so within that bolt circle, this gasket is
placed. So, here we have different dimensions like D, hG that you can observe from this figure.
So, D is basically mean diameter of gasket, C you can observe from this which is equal to
1
𝐹𝐵 ℎ𝐺 2
(0.31 + 190 𝑝𝐷 3 ) , so here FB is basically bolt load in kgf, hG is 0.5, bolt circle diameter minus

D and half of that will be equal to hG. So, in this way, you can find values of C as well as D for
different connections and whatever connection is present in your case, you can compute the
thickness of flat head for that case, okay.

So, in this way, we design the flat heads, okay.

(Refer Slide Time: 14:23)

Next head I am having is the conical head. Now geometry of the cone may be compared with that
of cylinder in which the diameter is continuously changing, okay? So, conical shape if I am
considering, it is like this, so diameter is continuously changing from top to bottom. This suggest
that cylindrical shell equation can also be used in this case provided the diameter is appropriately
modified incorporating the conical shape of the cylinder.
So, as this diameter is continuously changing, we usually design conical head in two ways, first is
we compute thickness of head at the junction, now what is junction. Junction is basically the part
where it is attached to the shell. For example, if I am having this conical shape, you can see this
image where the conical shape is there, you see, this is the conical shape which is attached to the
shell. So, if the curved part is not there, it will be directly attached to the shell.

So, at the junction means, at this part where it is attached to the shell. For example, if I am having
this curved section, then this curved section will be part of conical head and somewhere here it
will be attached to the shell. So, in that case, at the junction means, wherever this curvature occurs,
and if that curvature will not be there, it is directly attached to the shell, so at this point we compute
the thickness of conical head at the junction, so at the junction means this part, whether it is
attached to the shell over here or not, okay.

Thickness of this part we will compute as that at the junction. So, one thickness is at the junction,
second thickness is away from the junction as diameter is continuously changing, okay. So, we
will compute thickness away from the junction, away from the junction means, to this side, okay,
if I need to calculate the thickness for this section. So that will be calculated away from the
junction, away from the junction means L distance away from the junction.

So, what is that L? How it will be computed? That all we will discuss. Now, if you see this
diagram, here a few more points I need to focus. If you consider this Ri, what is Ri that is the
radius of knuckle part. If knuckle part is there, because if you remember we have discussed in
conical head, that if apex angle exceeds 60o, it means the conical section should have curved or
knuckle part, so that knuckle part must have the radius Ri okay.

So that you can see in this diagram. De if you see, D is basically outer diameter of head and here
we have Dk, Dk is basically the diameter where I am computing the thickness and α is half of apex
angle. Usually this section is defined as apex angle, so α is half of apex angle and in this case,
angle size is also given. So, for this case, ψ is equal to α. So, why these 2 terminologies are used
for same purpose because there are some cases where ψ is not equal to α. We will discuss that
case also. So, to distinguish the two cases, α and ψ are two different angles which are used in
conical heads.
(Refer Slide Time: 18:08)

So, as far as design is concerned, first we will start design from the junction, so we will compute
thickness of head at the junction. At the junction, you have understood, it means we are computing
thickness at this section. So thickness of cylindrical and conical section within length L from the
𝑃𝐷 𝐶
𝑜
junction shall be determined as = 200𝑓𝐽 . This expression we will use and I guess you understand

why this 200, because f and p are in different units, p should be in kgf/cm2 and f should be in
kgf/mm2.

So, this is expression is available in quote, where J is the joint efficiency factor, D is the outer
diameter of conical section as you can find from this image, and C is the factor taking into account
stress in the knuckle. So, according to these knuckles, factor C varies. That we will discuss how
𝐷 𝑡
𝑒
it varies and L is the distance from the junction which can be found as 0.5√𝐶𝑜𝑠𝜓 . So, this

expression is used to calculate L where this t is basically the actual thickness of the head at the
junction.

And then considering this thickness at the junction, we will compute distance L to calculate
thickness L distance away from the junction. I guess you are understanding this.
(Refer Slide Time: 19:56)
So, here you see, ψ I am having, ψ is the difference between angle of slope of two adjoining conical
sections and I have already discussed why the ψ is there. ψ and α both angles are there, but in
some cases these two angles are different. Therefore, we have used different notation for these
two. How it looks different, that we will discuss.

Next parameter I am having is Ri. Ri is basically inside radius of transition knuckle which shall
be taken as 0.01Dk. In case of conical section without knuckle transition in mm. So, if you see
this Ri, Ri basically denotes the knuckle transition part and if this is knuckle transition part is not
there, then Ri will be equal to 0.01Dk, okay. So that value I need to consider while designing.

What is Dk. It is the inside diameter of conical section or end at the position under consideration.
So, wherever I am finding the thickness, diameter at that particular section will be denoted as Dk.
For example, if I am finding at the junction, Dk will be equal to Di, inner diameter at that particular
section.
(Refer Slide Time: 21:21)
Now, in this slide, we will see how to find out value of C as a function of ψ and this is not Ri, this
is Ri, Ri /De, okay. This is stable in code IS: 2825, you can find this table as 3.6. So, now if you
see this stable, here in the first column I am having different angles of ψ and here we have Ri/De,
𝑅𝑖
here we have different values of and all these values you see are nothing but values of C.
𝐷𝑒

So, if I am not having different values of ψ as well as α, I can consider ψ equal to α. If it is

different, that difference will be mentioned okay? Now, we have two conditions that when knuckle
part is there and when knuckle part is not there, so when knuckle part is not there, it means Ri
would be equal to 0.01Dk and Dk, if I am calculating thickness at the junction, Dk should be Di,
𝑅
okay. If you see this table, here we have 𝐷𝑖 and here De, if you remember, that is nothing but the
𝑒

outer diameter and for calculation purpose we can assume Di should be equal to Do. It means Dk
𝑅
should be equal to Do, it means 𝐷 𝑖 equal to 0.01.
𝑘

So, what is the conclusion? If I am not using the knuckle part, it means C value you have to find
through this column only corresponding to different ψ value. For example, ψ is given as 45 and
in this problem it is mentioned that knuckle part is not given, so you can this table as corresponding
𝑅
to 45 and 𝐷𝑖 0.01 value available as 2.05, so that 2.05 you can consider value of C. However, if
𝑒

𝑅
I am having knuckle part, 𝐷 𝑖 will always be greater than 0.01.
𝑘
 𝑅𝑖
So considering the value of you can understand whether knuckle part is there or knuckle part
𝐷𝑘

is not there and this point will be more clear when we will solve a few examples based on this.
Now we have thickness of conical part away from the junction, as we have discussed away from
the junction means L distance away from the junction.
(Refer Slide Time: 24:04)

So to calculate thickness of that part, we have this expression as

𝑃𝐷𝑘 1
𝑡 = 200𝑓𝐽−𝑝 ∗ 𝐶𝑜𝑠𝛼. Here, p is the design pressure, why 200 is there, I hope you have idea this till

now, because p and f are different units, J is the joint efficiency factor.

So if you see this expression, what is Dk, Dk you can see from this diagram that where I am finding
out the thickness, inner diameter of that part will be called as Dk, okay. So, if you consider this
particular section, this section we can use to find out value of Dk, okay, as it is shown in this
diagram more clearly. So, for example, this is Dk and here we have A, B, C and E and F, D I am
not taking over here as Dk and Di are already included alphabet D, so here we need to find out
expression of Dk, okay.

Now if you see, this is basically Dk and to compute Dk, I have to focus on this part, okay. Now,
further, corresponding to this, I can draw a line over here and I can focus on this particular section.
So, if you consider this section, as well as this section, we can apply equal angle theory here. So,
𝐶𝐸
here, sinα if I need to calculate, sinα would be 𝐵𝐸 and BE is nothing but length L. So, CE how
 𝐹𝐸−𝐹𝐶
you can compute, , using this geometry you can find out, I think that you can make, it is not
𝐵𝐸

very much complicated.

𝐷𝑖 𝐷𝑘
FE is basically and FC is that you can see from the geometry and this B is nothing but the
2 2

L. Then further I can replace Di with this where the ts is the standard thickness of shell because it
is available at this particular section where conical section has equal inside diameter to shell. On
further resolving this expression, you can find out the final expression for Dk, so Dk you can find
out this way. So, here we have conical section and we have seen design of thickness at the junction
and away from the junction.
(Refer Slide Time: 26:46)

In this slide, you can focus on this particular image, here we have angle α and ψ, so both ψ is equal
to α in this particular figure. However, if I am focusing on this, here we have this α1, here α2 and
ψ, so here basically ψ as well as this α, these angles are different. So, if conical section is placed
in this way, we can have different value of ψ and α. So, this ψ and α may be same or may not be
same, so that depends on how conical section is made okay. So, that was about the conical section.
(Refer Slide Time: 27:41)
Now we will discuss design of torispherical heads as well as ellipsoidal heads. So, these heads are
basically coming under dished or formed heads. So, you see this image, where this figure is of
torispherical head and this figure is of ellipsoidal heads, so both heads are basically dished heads,
however, in ellipsoidal heads, we have more curved part in comparison to torispherical heads and
these two heads are designed using same expression as these are dished heads.

Thickness of dished heads can be determined by following expression which is

𝑃𝐷 𝐶
𝑜
𝑡 = 200𝑓𝐽 , here again I am having 200, I guess you understand, I do not need to repeat that again

and again. So, here you see, the main point to be considered is how to find C. C is basically the
shape factor, and how we can find the factor C, that is the main concern over here. Other
parameters how you need to take, that you are already aware. So let us discuss how to find value
of C.
(Refer Slide Time: 28:56)
Now, if you consider this particular slide, here I am having this image, if you see this, which is
figure 3.7 in IS: 2825, so that figure you can refer, so Y axis of this figure is having factor C
ℎ𝑒 ℎ𝑒
whereas X-axis has value of . So, once I will calculate , I can find value of C and if you
𝐷𝑜 𝐷𝑜
𝑡
observe this graph, here I am having solid lines for 𝐷 when it is varying from 0.002 to 0.004 and
𝑜

here also I am having broken lines and these lines again vary from 0.5 to 5, okay. On this axis, we
𝐷
have . So this image I have to use to compute the value of C.
√𝑡𝐷𝑜

ℎ
So, first of all, to use this, we have to find 𝐷𝑒 Do is the outer diameter of head or shell, so that we
𝑜

already know, how to find out the value of he, that I have to speak on. So, he is basically effective
𝐷𝑜2 𝐷𝑜 𝑟𝑜
outside height of the end and that can be determined as, here we have he, and √ , so to
4𝑅𝑜 2

compute he, we need to find ho, we need to find this expression and whichever is minimum, that I
need to take as he value, and whichever is minimum among these there, that I need to take as value
of he.

Now, why we are taking minimum among these three as he, because when I am having minimum
he, he is basically, if you see the diagram, if you focus on this particular slide, he is basically this
height. he is basically this he o that is effective height. So, if I am having more value of he, it
means it has more curved part and it will include higher manufacturing cost. Therefore, if I am
having three different values to find out he, I will select the minimum one.
Now, among these three values, ho you can find out using this expression where Ro, Do/2 and small
ro, all these values you know about the head, and similarly this value I will find, and I will choose
the minimum value among these three as he. So, we have seen that how he will be computed.
(Refer Slide Time: 32:14)

And once he will be known, in this figure, you can make a line, okay. Let us say he is 0.2, so
corresponding to 0.2 you can make a line and how you need to find out value of C is, when ends
𝑡 𝑡
are fully compensated, read C from full curves that is 𝐷 0.002 to 0.00 that is 𝐷 equal to 0.002 to
𝑜 𝑜

𝑡
0.04 and we can make interpolation also.
𝐷𝑜

𝑜 𝑃𝐷 𝐶
So, how you need to compute this, if you see the expression, 𝑡 = 200𝑓𝐽 . Now, here you need to
𝑡
find C, so how I will further consider is 𝐷 because if it is fully compensated, we need to see value
𝑜

𝑡 𝑡
, so 𝐷 I will extract and if you consider this expression, okay, this whole expression will be a
𝐷𝑜 𝑜

ℎ𝑒
constant. Let us say that value is X*C. So, corresponding to 0.2, you need to choose value of
𝐷𝑜
𝑡
and corresponding value of C in such a way so that this expression should be satisfied. So, I
𝐷𝑜

hope you are getting how to find out value of C and it will be more clear when we solve a few
examples based on that.
There I will elaborate this in more detail. Now, when end remains uncompensated, read C from
𝐷
broken line that is = 0.5 𝑡𝑜 5, it is varying as you can see from these broken lines. So what
√𝑡𝐷𝑜

is this compensation and un-compensation, that we will cover in lecture 4 and 5 of this week? So,
basically when we make any opening, that opening may be compensated or may not be
compensated, so based on that, we consider design of heads.

So, till now, you please bear compensation and uncompensation as it is, I will elaborate that when
time comes, okay. So, if you consider this statement, here small d is diameter of largest
uncompensated opening. For example, in head I am preparing two, three openings. One is having,
let us say diameter 0.05 metre and another is having let us say 0.7 m, so d would be 0.7 in that
ℎ𝑒
case. So, considering that d, you need to find out value of C for given value. I hope it is clear.
𝐷𝑜

ℎ
Further, values of C corresponding to 𝐷𝑒 you can find from table 4.1A as well s B from book and
𝑜

this book is a book from B. C. Bhattacharya, so that book you can use.
(Refer Slide Time: 35:27)

So this is basically table 4A, which is used for without opening or with fully compensated openings
𝑡
and here I am having different values of which is varying from 0.002 to 0.04 as it is appearing
𝐷𝑜
ℎ𝑒
in the graph, and here I am having different value of and all these values are values of C,
𝐷𝑜

okay.
(Refer Slide Time: 35:53)

𝐷
And this table I am having is for uncompensated opening or that is table 4.1B where is shown
√𝑡𝐷𝑜
ℎ𝑒
from 0.5 to 5 and corresponding to , you can read different values of C from this table. So, in
𝐷𝑜

this way, you can use that graph as well as this table to find out value of C for torispherical as well
as ellipsoidal heads. So, as I have discussed that torispherical head as well as ellipsoidal head,
both heads are designed using same expression, so what is the difference between these two.

Difference will lie when I am having major and minor axis, okay. So, when I am considering
ℎ𝑒
ellipsoidal head, it has specific major and minor axis, and accordingly value will vary.
𝐷𝑜
ℎ
However, that major and minor axis ratio is not maintained in torispherical head. So, based on 𝐷𝑒
𝑜

you can find different value of C for ellipsoidal heads as well as for torispherical head, so in that
way you can make different designs for these two heads.

Now, here we will add 6% extra in t for reduction in thickness as toris section. So, when I am
preparing the forming section, some material is lost, therefore, we consider additional material
while designing that forming section, so if I am considering torispherical head, ellipsoidal head,
conical head with knuckle, or flat formed head, wherever I am having that formed section, I will
add 6% extra to the minimum thickness. That guideline you need to follow, okay.
(Refer Slide Time: 37:46)
So, once I am having this, we will calculate the blank diameter. What is blank diameter? It is the
diameter of plate with which form Ls can be fabricated. So, if I am having this ellipsoidal head,
the diameter means the diameter of this whole section, whole plate through which this ellipsoidal
𝐷 2
head is prepared, so that would be equal to 𝐷𝑜 + 42𝑜 + 3   𝑟𝑖 + 2𝑆𝑓 . Sf you can understand, Sf is

basically this straight part, which is beyond the curved section.

So, for t is less than 25 mm, we will use this expression, otherwise we will use this expression,
okay. So, based on these, we can find out blank diameter. And here I am stopping lecture 2 of
week 2. I will continue design of heads in next lecture where I will illustrate a few examples for
designer heads. And that is all for now, thank you.